Method for qualitatively identifying oxidation stability of atorvastatin calcium bulk drug by adopting multispectral fusion technology

Abstract

The invention discloses a method for qualitatively identifying oxidation stability of an atorvastatin calcium bulk drug by adopting a multispectral fusion technology, and belongs to the technical field of detection analysis, and the method comprises the following steps: S1, preparing amorphous atorvastatin calcium; s2, grinding, screening, mixing and tabletting the atorvastatin calcium raw material medicine with the crystal form I and the amorphous atorvastatin calcium; s3, detecting the tabletting sample to obtain near infrared spectrum data and Raman spectrum data; s4, performing feature extraction on the near infrared spectrum data and the Raman spectrum data, fusing near infrared spectrum features and Raman spectrum features, and establishing a mathematical model of the fused features after standardization treatment and amorphous atorvastatin calcium content data in the mixed sample; and S5, establishing classification boundaries of oxidability and oxidation stability by using a plurality of atorvastatin calcium samples with known oxidation stability, and identifying the oxidation stability of the unknown atorvastatin calcium sample to be detected according to the classification boundaries by using the mathematical model established in S4.

Description

Technical Field

[0001] The present invention belongs to the technical field of detection and analysis, and particularly relates to a method for qualitatively identifying the oxidative stability of atorvastatin calcium raw materials by using multi-spectral fusion technology. Background Art

[0002] Atorvastatin calcium is commonly used to treat hypercholesterolemia. It is a selective, competitive inhibitor of hydroxymethylglutaryl coenzyme A (HMG-CoA) reductase, capable of lowering plasma cholesterol and lipoprotein levels. It has the advantages of strong lipid-lowering effects and minimal adverse reactions, and is now widely used clinically. The presence of a pyrrole ring structure in the atorvastatin calcium molecule makes it susceptible to oxidative degradation under conditions of light, high temperature, or high humidity, generating a variety of oxidative impurities, including epoxides and ring-opened derivatives. These oxidative impurities not only reduce the drug's potency but also may produce potential toxic side effects.

[0003] Accurate evaluation of the oxidative stability of APIs is a key step in ensuring the quality of preparations and drug safety. Current testing methods mainly rely on high-performance liquid chromatography combined with accelerated stability testing for oxidative degradation analysis, but have the following shortcomings: (1) Destructive testing results in the inability to reuse samples; (2) Forced degradation tests lasting several weeks are required, which is time-consuming (stable for 3 months at 40°C ± 2°C and 75% ± 2% relative humidity).

[0004] In the prior art, researchers have developed emerging methods for drug stability testing, such as artificial intelligence-assisted stability prediction models and microreactor-based oxidative environment simulations, to assess the degradation behavior of APIs and formulations. For example, Chinese patent publication CN107255646A discloses a method for rapid and quantitative prediction of drug stability. This invention uses XRD patterns combined with chemometric methods to establish a model for the maximum single impurity and total impurity after stabilization of drug crystals. This model is used to predict drug stability, allowing for rapid assessment of the impact of different experimental conditions on drug stability. Chinese patent publication CN119763043A discloses a drug stability monitoring method and system based on image analysis. The method comprises: obtaining multiple first images of the drug to be monitored; the multiple first images are captured under identical lighting conditions and within a predetermined capture period; extracting color, shape, and texture features from each first image; and obtaining stability monitoring results for the drug to be monitored based on the color, shape, and texture features. However, none of these methods are suitable for qualitative testing of the oxidative stability of atorvastatin calcium API.

[0005] In view of this, there is an urgent need to develop a method for qualitatively identifying the oxidative stability of atorvastatin calcium raw materials, so as to non-destructively identify the oxidative stability of samples in a relatively short period of time, and provide a new idea for the rapid and non-destructive identification of the oxidative stability of drugs. Summary of the Invention

[0006] The present invention provides a method for qualitatively identifying the oxidative stability of atorvastatin calcium raw material pharmaceutical by using multi-spectral fusion technology. The method can improve the ability to identify the quality of the raw material pharmaceutical, thereby ensuring the quality and safety of the pharmaceutical during the production process.

[0007] The specific technical solutions adopted are as follows:

[0008] A method for qualitatively identifying the oxidative stability of atorvastatin calcium raw material using multispectral fusion technology comprises the following steps:

[0009] S1: Preparation of amorphous atorvastatin calcium;

[0010] S2: grinding, sieving and mixing the atorvastatin calcium crystalline form I and the amorphous atorvastatin calcium prepared in S1, and then tableting the mixed mixture;

[0011] S3: Using near-infrared spectroscopy and Raman spectroscopy to detect the compressed sample, and obtaining near-infrared spectral data and Raman spectral data;

[0012] S4: extracting features from the near-infrared spectral data and the Raman spectral data, fusing the extracted near-infrared spectral features and the Raman spectral features, and establishing a mathematical model for the standardized fusion features and the amorphous atorvastatin calcium content data in the mixed sample;

[0013] S5: Using multiple atorvastatin calcium samples with known oxidative stability, establish classification boundaries for oxidizability and oxidative stability, and use the mathematical model established in S4 to identify the oxidative stability of the unknown atorvastatin calcium sample to be tested based on the classification boundaries.

[0014] This method combines near-infrared spectroscopy and Raman spectroscopy to acquire multidimensional complementary information about chemical bond vibrational modes by capturing the molecular vibrational signatures of a sample. Using a feature-layer fusion algorithm, the two spectral data are fused to construct a multidimensional fusion correction model based on chemometrics, enabling qualitative identification of the sample's oxidative stability. This method transcends the dimensional limitations of single spectral techniques, offers non-destructive testing capabilities, and can meet the demand for rapid qualitative identification of a sample's oxidative stability.

[0015] Preferably, in step S1, the crystalline form I atorvastatin calcium is dissolved in a mixed solvent of acetone and ethyl acetate, stirred until completely dissolved, and then distilled under reduced pressure to reduce the volume of the solution, and then diethyl ether is added at room temperature, and distilled under reduced pressure again to obtain amorphous atorvastatin calcium.

[0016] Further preferably, the mass volume ratio of Form I atorvastatin calcium to the mixed solvent is 1 g:10-20 mL; and the volume ratio of acetone to ethyl acetate in the mixed solvent is 1:1-5.

[0017] More preferably, the temperature of the first vacuum distillation is 30-50°C, and the temperature of the second vacuum distillation is 40-50°C.

[0018] In step S2, amorphous atorvastatin calcium and atorvastatin calcium API of the same particle size range are mixed according to different preset mass fractions.

[0019] Preferably, the particle sizes of amorphous atorvastatin calcium and crystalline form I atorvastatin calcium API are both in the range of 100-150 mesh, and the amorphous atorvastatin calcium and crystalline form I atorvastatin calcium API are mixed in a mass ratio of 1:0.05-1.

[0020] Preferably, the diameter of the tablet sample obtained after tableting is 6-10 mm.

[0021] Preferably, in step S3, when the near-infrared spectroscopy technique is used for detection, the near-infrared spectrometer used is equipped with a dual-polarization fiber optic reflection probe. During the detection process, the number of scans is 16-64 times, and the resolution is 8 cm -1 , spectral range 4000–12000 cm -1 When using Raman spectroscopy for detection, the laser wavelength is 785nm, the number of integration times is 2-10 times, the integration time is 5-15s, the laser intensity is set to 5%-15%, and the Raman shift resolution is an average of 1.63cm -1 , Raman shift range is 40-3370cm -1 .

[0022] Preferably, in step S4, the near-infrared spectral data and the Raman spectral data are preprocessed using at least one of the Savitzky-Golay first-order derivative method and the standard normal variable transformation method, and then the near-infrared spectral features and the Raman spectral features are extracted using the competitive adaptive reweighted sampling method.

[0023] Preferably, in step S4, a mathematical model is established using partial least squares method, in which the fused features are used as independent variables, and the content of amorphous atorvastatin calcium in the mixed tablet sample is used as the dependent variable.

[0024] Furthermore, in step S5, for a plurality of atorvastatin calcium samples (including stable samples and easily oxidized samples) with known oxidative stability, the content of amorphous atorvastatin calcium therein is obtained using a mathematical model, and the maximum value of the content of amorphous atorvastatin calcium in the corresponding stable samples is calculated as the classification boundary. If the calculated value of the amorphous atorvastatin calcium content in the unknown atorvastatin calcium sample to be tested is greater than the classification boundary, it is determined to be easily oxidizable, otherwise it is determined to be oxidatively stable.

[0025] Compared with the prior art, the present invention has the following beneficial effects:

[0026] The present invention provides a method for qualitatively identifying the oxidative stability of atorvastatin calcium API using multispectral fusion technology. This method integrates multidimensional spectral information to achieve rapid and non-destructive identification of the oxidative stability of atorvastatin calcium API, providing a basis for quality control of atorvastatin calcium API. It also provides new ideas for the quality identification of other APIs, thereby helping to improve drug quality. BRIEF DESCRIPTION OF THE DRAWINGS

[0027] Figure 1 Near-infrared and Raman spectra of the mixed sample.

[0028] Figure 2 This is a graph showing how the cross-validated root mean square error and the number of selected features change with the number of Monte Carlo sampling times during the near-infrared spectral feature extraction process in Example 1.

[0029] Figure 3 This is a graph showing how the cross-validated root mean square error and the number of selected features change with the number of Monte Carlo sampling times during the Raman spectrum feature extraction process in Example 1.

[0030] Figure 4 This is a graph showing how the cross-validated root mean square error and the number of selected features change with the number of Monte Carlo sampling times during the near-infrared spectral feature extraction process in Example 2.

[0031] Figure 5 This is a graph showing how the cross-validated root mean square error and the number of selected features change with the number of Monte Carlo sampling times during the Raman spectrum feature extraction process in Example 2. DETAILED DESCRIPTION

[0032] The present invention will be further illustrated below with reference to the examples. It should be understood that these examples are intended to illustrate the present invention only and are not intended to limit the scope of the invention. The operating methods in the following examples where specific conditions are not specified are generally performed under conventional conditions or as recommended by the manufacturer. The contents not described in detail in this specification belong to the prior art known to those skilled in the art.

[0033] Example 1

[0034] In this embodiment, a method for qualitatively identifying the oxidative stability of atorvastatin calcium raw material drug using near-infrared spectroscopy and Raman spectroscopy fusion technology is described, and the specific steps are as follows:

[0035] S1: Preparation of amorphous atorvastatin calcium.

[0036] Dissolve 20 g of Form I atorvastatin calcium in 200 mL of a mixed solvent of acetone and ethyl acetate (acetone:ethyl acetate = 1:1, v / v) and stir until completely dissolved to obtain a clear, transparent solution. Distill under reduced pressure at 50°C until the solution volume is approximately 60 mL. Cool the solution to room temperature, add 60 mL of anhydrous ether, and distill under reduced pressure at 50°C again for 2 h to obtain amorphous atorvastatin calcium.

[0037] S2: Grind, sieve and mix the crystal form I atorvastatin calcium bulk drug and the amorphous atorvastatin calcium prepared in S1, and then perform tableting after mixing.

[0038] The atorvastatin calcium bulk drug substance in form I and the amorphous atorvastatin calcium prepared in the above steps were ground using an agate mortar and sieved after appropriate grinding to obtain particles with a particle size range of 0.100 mm-0.154 mm (100 mesh-150 mesh).

[0039] Atorvastatin calcium Form I API and amorphous atorvastatin calcium were weighed at amorphous atorvastatin calcium contents of 0, 5, 10, 15, 20, and 30% and placed in polytetrafluoroethylene tubes, numbered as samples h0, h1, h2, h3, h4, and h5, respectively. The tubes were placed in a homogenizer and removed every 30 minutes and placed on an oscillator for at least 3 hours. After mixing, approximately 53 mg of sample was weighed, placed in a tableting mold, and pressed into tablets (8 mm diameter).

[0040] S3: Use near-infrared spectroscopy and Raman spectroscopy to detect the compressed samples and obtain near-infrared spectroscopy data and Raman spectroscopy data.

[0041] The compressed samples prepared in the above steps were tested using a near-infrared spectrometer with a dual-polarization fiber optic reflectance probe (ABB TALYS, ABB Beverage Co., Ltd.). During the test, the number of scans was 32 and the resolution was 8 cm. -1 ; Spectral range is 4000-12000cm -1 , obtain near-infrared spectral data, remove the overloaded part of the front section of the spectrum and the information-free part of the back section, the results are as follows Figure 1 shown.

[0042] The pressed tablets prepared in the above steps were tested using a Raman spectrometer (BWS475-785S, B&W Tech Optoelectronics (Shanghai) Co., Ltd.). During the test, the laser wavelength was 785 nm, the number of integrations was 3, the integration time was 10 s, the laser intensity was set to 10%, and the average Raman shift resolution was 1.63 cm. -1 , Raman shift range is 40-3370cm -1 , obtain Raman spectrum data, remove the overloaded part of the front section of the spectrum and the information-free part of the back section, the results are as follows Figure 1 shown.

[0043] S4: Extract features from the near-infrared spectral data and the Raman spectral data, fuse the extracted near-infrared spectral features and the Raman spectral features, and establish a mathematical model of the standardized fusion features and the amorphous atorvastatin calcium content data in the mixed sample.

[0044] After preprocessing the near-infrared spectral data and Raman spectral data respectively using the standard normal variable transformation method, the features of the near-infrared spectral data and Raman spectral data were extracted using the competitive adaptive reweighted sampling method to obtain the near-infrared spectral features and Raman spectral features, and then the near-infrared spectral features and Raman spectral features were extracted using the competitive adaptive reweighted sampling method.

[0045] In the process of competitive adaptive reweighted sampling, the number of Monte Carlo sampling was set to 50, and the competitive adaptive reweighted sampling method was repeated 500 times. The feature corresponding to the minimum root mean square error was selected as the target feature to be extracted. The extracted near-infrared spectral features and Raman spectral features were wavenumbers 6542.06, 8216.15, 8208.44, and 7220.96 cm, respectively. -1 and wave numbers 1087.87, 860.68, 1021.20 cm -1 . Figure 2 and Figure 3 The cross-validation root mean square error and the number of selected features changing with the Monte Carlo sampling times during the near-infrared spectral feature extraction and Raman spectral feature extraction are respectively shown.

[0046] The near-infrared spectral feature and Raman spectral feature data were fused and standardized, and the partial least squares method was used to establish a mathematical model M. In this mathematical model, the fused features were used as independent variables, and the content of amorphous atorvastatin calcium in the mixed tablet sample was used as the dependent variable. In this mathematical model, the number of principal components of the model was 1, and R 2 It is 0.9801, indicating that the model fits well and can explain most of the variation.

[0047] S5: Using multiple atorvastatin calcium samples with known oxidative stability, establish classification boundaries for oxidizability and oxidative stability, and use the mathematical model established in S4 to identify the oxidative stability of the unknown atorvastatin calcium sample to be tested based on the classification boundaries.

[0048] For multiple atorvastatin calcium samples with known oxidative stability (including stable samples and easily oxidized samples), the same grinding, screening and tableting operations as the above steps were performed, and their spectral data were obtained using near-infrared spectroscopy detection technology and Raman spectroscopy technology. The same steps as in S4 were used to extract their near-infrared spectral features and Raman spectral features, and after fusion and standardization, the content of amorphous atorvastatin calcium was obtained using mathematical model M. The results are shown in Table 1, where the samples with labels beginning with "S" (stable) are known stable samples, and the samples with labels beginning with "US" (unstable) are known easily oxidized samples.

[0049] Table 1 Predicted values ​​of amorphous content

[0050]

[0051] The maximum value (21.3%) of the amorphous atorvastatin calcium content in the corresponding stable sample was calculated as the classification limit, and the mathematical model M was used to identify the oxidative stability of the unknown atorvastatin calcium sample to be tested. If the calculated value of the amorphous atorvastatin calcium content in the unknown atorvastatin calcium sample to be tested was greater than the classification limit, it was identified as easily oxidizable, otherwise it was identified as oxidatively stable.

[0052] Verify the accuracy A of the mathematical model M. The calculation formula is:

[0053]

[0054] Where m is the number of samples, and a is the number of misclassified samples in the sample.

[0055] Calculation shows that the model accuracy of the above mathematical model M for classifying whether the sample is stable is 1.0 (for atorvastatin calcium samples with known oxidative stability), indicating that the mathematical model can better identify the oxidative stability of atorvastatin calcium raw materials.

[0056] Example 2

[0057] In this embodiment, a method for qualitatively identifying the oxidative stability of atorvastatin calcium raw material drug using near-infrared spectroscopy and Raman spectroscopy fusion technology is described, and the specific steps are as follows:

[0058] S1, S2, and S3 are the same as in Example 1.

[0059] S4: Extract features from the near-infrared spectral data and the Raman spectral data, fuse the extracted near-infrared spectral features and the Raman spectral features, and establish a mathematical model of the standardized fusion features and the amorphous atorvastatin calcium content data in the mixed sample.

[0060] After preprocessing the near-infrared spectral data and Raman spectral data using the Savitzky-Golay first-order derivative and standard normal variable transformation methods, respectively, the features of the near-infrared spectral data and Raman spectral data were extracted using the competitive adaptive reweighted sampling method to obtain the near-infrared spectral features and Raman spectral features, and then the near-infrared spectral features and Raman spectral features were extracted using the competitive adaptive reweighted sampling method.

[0061] In the process of competitive adaptive reweighted sampling method, the number of Monte Carlo sampling was set to 50, and the competitive adaptive reweighted sampling method was repeated 500 times. The feature corresponding to the minimum root mean square error in 500 repetitions was selected as the target feature to be extracted. The extracted near-infrared spectral features and Raman spectral features were wavenumbers 8223.87, 7922.99, 7853.56, 8247.01, 6819.79, 6950.94, 7722.41, 7730.13, 7868.99, 7745.56, 8231.58, 7614.41, 7876.71, 8216.15, 7760.99, 7058.95, 9157.35, 7930.71, 8108.15, and 644 1.77, 7568.12, 8393.59, 9149.63, 7691.55, 6495.78, 7814.99, 7660.69, 7799.56, 6966.37, 6488.06, 5685.73, 6503.49, 8092.72, 7699.27, 7645.26, 9134.20, 8239.30, 5678.02, 7066.66, 6526.63, 6958.66, 9141. 92, 6395.48, 7953.85, 7946.14, 6464.92, 7992.43, 7622.12, 9118.77, 8139.01, 7074.38, 9172.78, 6518.92, 6788.93, 7884.42, 6233.48, 6511.20, 8208.44, 8532.46, 6534.35, 7529.54, 7089.81, 5971.18, 8293.30 , 8493.88, 8185.29, 8625.03, 8486.17, 8200.72, 6889.22, 5963.46, 8632.75, 8640.46, 8432.16, 8424.45, 8409.02, 7490.97, 8308.73, 8355.02, 8324.16, 7190.10, 5446.58, 8956.76, 5508.29, 5076.27, 7915.28cm -1 and wave numbers 1085.92, 860.68cm -1 . Figure 4 and Figure 5 The cross-validation root mean square error and the number of selected features varying with the number of Monte Carlo sampling times during near-infrared and Raman spectral feature extraction are shown respectively.

[0062] The near-infrared spectral feature and Raman spectral feature data were fused and standardized, and the partial least squares method was used to establish a mathematical model M. In this mathematical model, the fused features were used as independent variables, and the content of amorphous atorvastatin calcium in the mixed tablet sample was used as the dependent variable. In this mathematical model, the number of principal components of the model was 1, and R 2 It is 0.9976, indicating that the model fits well and can explain most of the variation.

[0063] S5: Using multiple atorvastatin calcium samples with known oxidative stability, establish classification boundaries for oxidizability and oxidative stability, and use the mathematical model established in S4 to identify the oxidative stability of the unknown atorvastatin calcium sample to be tested based on the classification boundaries.

[0064] For multiple atorvastatin calcium samples with known oxidative stability (including stable samples and easily oxidized samples), the same grinding, screening and tableting operations as the above steps were performed, and their spectral data were obtained using near-infrared spectroscopy detection technology and Raman spectroscopy technology. The same steps as in S4 were used to extract their near-infrared spectral features and Raman spectral features, and after fusion and standardization, the content of amorphous atorvastatin calcium therein was obtained using the mathematical model M. The calculation results are shown in Table 2.

[0065] Table 2 Predicted values ​​of amorphous content

[0066]

[0067] The maximum value (8.73%) of the amorphous atorvastatin calcium content in the corresponding stable sample was calculated as the classification limit, and the mathematical model M was used to identify the oxidative stability of the unknown atorvastatin calcium sample to be tested. If the calculated value of the amorphous atorvastatin calcium content in the unknown atorvastatin calcium sample to be tested was greater than the classification limit, it was identified as easily oxidizable, otherwise it was identified as oxidatively stable.

[0068] The model accuracy was calculated using the same method as in Example 1, and the model accuracy of the above mathematical model for classifying whether the sample is stable was 1.0 (for atorvastatin calcium samples with known oxidative stability), indicating that the model can better identify the oxidative stability of atorvastatin calcium raw materials.

[0069] The embodiments described above provide a detailed description of the technical solutions of the present invention. It should be understood that the above are only specific embodiments of the present invention and are not intended to limit the present invention. Any modifications, supplements or similar substitutions made within the scope of the principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for qualitatively identifying the oxidative stability of atorvastatin calcium raw material using multispectral fusion technology, characterized in that: The following steps are involved: S1: Preparation of amorphous atorvastatin calcium; S2: grinding, sieving and mixing the atorvastatin calcium crystalline form I and the amorphous atorvastatin calcium prepared in S1, and then tableting the mixed mixture; S3: Using near-infrared spectroscopy and Raman spectroscopy to detect the compressed sample, and obtaining near-infrared spectral data and Raman spectral data; S4: extracting features from the near-infrared spectral data and the Raman spectral data, fusing the extracted near-infrared spectral features and the Raman spectral features, and establishing a mathematical model for the standardized fusion features and the amorphous atorvastatin calcium content data in the mixed sample; S5: Using multiple atorvastatin calcium samples with known oxidative stability, establish classification boundaries for oxidizability and oxidative stability, and use the mathematical model established in S4 to identify the oxidative stability of the unknown atorvastatin calcium sample to be tested based on the classification boundaries.

2. The method for qualitatively identifying the oxidative stability of atorvastatin calcium bulk drug using multispectral fusion technology according to claim 1, characterized in that: In step S1, the crystal form I atorvastatin calcium is dissolved in a mixed solvent of acetone and ethyl acetate, stirred until completely dissolved, and then distilled under reduced pressure to reduce the volume of the solution. Subsequently, ether is added at room temperature and distilled under reduced pressure again to obtain amorphous atorvastatin calcium.

3. The method for qualitatively identifying the oxidative stability of atorvastatin calcium bulk drug using multispectral fusion technology according to claim 2, characterized in that: The mass volume ratio of Form I atorvastatin calcium to the mixed solvent is 1 g:10-20 mL; the volume ratio of acetone to ethyl acetate in the mixed solvent is 1:1-5.

4. The method for qualitatively identifying the oxidative stability of atorvastatin calcium bulk drug using multispectral fusion technology according to claim 1, characterized in that: In step S2, amorphous atorvastatin calcium and atorvastatin calcium API of the same particle size range are mixed according to different preset mass fractions.

5. The method for qualitatively identifying the oxidative stability of atorvastatin calcium bulk drug using multispectral fusion technology according to claim 4, characterized in that: The particle sizes of the amorphous atorvastatin calcium and the crystalline form atorvastatin calcium raw material are both within the range of 100 mesh to 150 mesh, and the amorphous atorvastatin calcium and the crystalline form atorvastatin calcium raw material are mixed in a mass ratio of 1:0.05-1.

6. The method for qualitatively identifying the oxidative stability of atorvastatin calcium bulk drug using multispectral fusion technology according to claim 1, characterized in that: In step S3, when using near-infrared spectroscopy technology for detection, the near-infrared spectrometer used is equipped with a dual-polarization fiber optic reflection probe. During the detection process, the number of scans is 16-64 times, and the resolution is 8 cm -1 , spectral range 4000–12000 cm -1 When using Raman spectroscopy for detection, the laser wavelength is 785nm, the number of integration times is 2-10 times, the integration time is 5-15s, the laser intensity is set to 5%-15%, and the Raman shift resolution is an average of 1.63cm -1 , Raman shift range is 40-3370cm -1 .

7. The method for qualitatively identifying the oxidative stability of atorvastatin calcium bulk drug using multispectral fusion technology according to claim 1, characterized in that: In step S4, the near-infrared spectral data and the Raman spectral data are preprocessed using at least one of the Savitzky-Golay first-order derivative method and the standard normal variable transformation method, and then the near-infrared spectral features and the Raman spectral features are extracted using the competitive adaptive reweighted sampling method.

8. The method for qualitatively identifying the oxidative stability of atorvastatin calcium bulk drug using multispectral fusion technology according to claim 1, characterized in that: In step S4, a mathematical model is established using the partial least squares method. In the mathematical model, the fused features serve as independent variables, and the content of amorphous atorvastatin calcium in the mixed tablet sample serves as the dependent variable.

9. The method for qualitatively identifying the oxidative stability of atorvastatin calcium bulk drug using multispectral fusion technology according to claim 1, characterized in that: In step S5, for a plurality of atorvastatin calcium samples with known oxidative stability, the content of amorphous atorvastatin calcium therein is obtained using a mathematical model, and the maximum value of the content of amorphous atorvastatin calcium in the corresponding stable sample is calculated as the classification limit. If the calculated value of the amorphous atorvastatin calcium content in the unknown atorvastatin calcium sample to be tested is greater than the classification limit, it is determined to be easily oxidizable, otherwise it is determined to be oxidatively stable.

Images